Dynamic image zero watermarking using Weber Local Descriptor and Contourlet Transform

At present, due to advancement in communication technology and internet, people are sharing images over social networking and other sites very efficiently but it has created problems regarding copyright protection and authentication of the images. In past 10-15 years, digital watermarking has been becoming a popular solution for these problems. Besides of being popular, it creates distortions in host image after watermark insertion. In many applications, this distortions are not tolerable. In such condition, Zero watermarking is a solution over normal watermarking where watermark is created using feature points of host image instead of adding into it. In this paper, we are proposing a dynamic zero watermarking algorithm based on Weber Local Descriptor and Contourlet Transform. Here zero-watermark is created using fusion of feature points detected by Weber Local Descriptor (WLD) in Contourlet Domain and the watermark. Fusion of WLD with Contourlet Transform leads to very robust feature point detection. We have adopted a dynamic approach, where the watermark image is adaptively resized according to number of feature points to create zero-watermark. This zero-watermark is adaptively scrambled using Arnold transform to improve its security. From Experimental results it is clear that our proposed algorithm possesses high robustness against various attacks and also gives better results over previous work in terms of recovery of watermark.